Poling issues weird stuff

Okay the upgrade went fine, go the custom pollers up and running for the Lieberts. But today the strangest thing started to happen, to the point I had to turn alerting off.

We had nodes show as down who are truly up. This is so frustrating, now what I know is the item is not down, when Orion reports it as down. Traffic on the network is not up, CPU utilization on the 4500 cores is around 60% off peak hours it is currently @ 40%. I don't think that we are having network issues, but I guess I could be. We have had reports of issues but once I check they are cleared up. That is within seconds of being notified.

In the past when I had issues the CPUs would peg at 100% for a short burst, we do not see heavy utilization on the switches or servers. We are getting no unusual hits on the access lists - deny statements. IOS's are at the latest relaease for each of the switches.

I have decreased the polling intervals, and increased the timeout of the reporting of the node being down. I have had 100s systems report down then up all throughout the day and night. It is very frustrating!!!! Somebody toss me some ideas here on other items to look at.

TIA, Stephen